Japanese scientists say they have broken through a technical barrier to developing smaller, more powerful computer microprocessors, a move that could guarantee companies will be able to improve semiconductor performance well into the 21st Century.
A coalition of Kyoto University, Fujitsu Laboratories Ltd. and Japan Science and Technology Corp.–a quasi-government agency–has developed a process that will let manufacturers cut smaller and smaller grooves into the silicon wafers that are used to produce chips.
But skeptics considered unlikely claims by the Japanese coalition that it represented the first in a series of breakthroughs that would advance chip designs by at least 10 years.
The developers of the technology–known as “doping” within the industry–indicated that the technology would be available within four years, according to an article in Electronic Engineering Times, a weekly trade publication. They said it would be the cornerstone of a series of technology advances that would reduce chip sizes by almost 300 percent while increasing the performance of some semiconductors by as much as 64 times.
Conventional wisdom indicates the industry won’t approach that kind of performance until 2010 at the earliest. Such an acceleration in the development cycle would explode Moore’s Law–an industrywide maxim that says the performance of semiconductors will double in performance roughly every 18 months.
But in the article, developers of the technology indicated that memory chips built using the smallest process possible could ship by the year 2002.
“I hope (the technology) will contribute to proving the Moore’s Law prophecy wrong, in a good direction,” Kyoto University professor Isao Yamada, who heads the Ion Beam Engineering Experimental Laboratory, was quoted as saying in the article.
The scientists’ claim met with widespread disbelief from analysts.
“If it were possible, that would be a very significant advance,” said Lynley Gwenapp, editor of the Microprocessor Report, a San Jose-based trade newsletter.
More Top Picks Best Smoke Detectors
But “this is only one process out of about 20 that are needed to advance the technology overall,” Gwenapp said. “It’s extremely unlikely that everything else that’s needed is going to be here by then.”
By itself, the Japanese advance will do little to move the industry to such a dramatically smaller process technology so much faster, said Rob Wilson, the West Coast bureau chief for Elecrtonic Engineering Times.
“There is so much more that has to be developed in order to build transistors at that size that you wonder what the Japanese are doing,” Wilson said. “The question (that) needs to be asked is, (do) they intend to bring the other technologies needed to support that process to market within four years as well?”
The industry is working at improving the technologies that will support such advances in chip designs, including using copper instead of aluminum for the on-chip wiring, better photographic techniques, and most important, new ways of etching ever-smaller designs on the semiconductor itself.
“All of these processes have to come together at the same time for them to be of any use,” said Dan Hutcheson, president of San Jose-based market research firm VLSI Research Inc.